Head to head on 20 boards: OpenAI places best on 16, DeepSeek places best on 3 and Nvidia places best on 1. All 3 ranked on 10 of 27 boards.
Two at a time on this screen. Tap one to swap it in.
| Board | Read |  Nvidia |  OpenAI |  DeepSeek |
|---|
| Chat2 | | | | |
|---|
 MultiChallengeScoretodaytoday | Nov 9todaytoday | | #182.0%best of the pickedGPT-6.1 Sol | #574.9%DeepSeek V4P1 Flash |
 Text ArenaArena ratingtodaytoday | Nov 9todaytoday | | #191,485best of the pickedGPT-5.6 Sol | #321,475DeepSeek V4.1 Flash |
| Reasoning7 | | | | |
|---|
 Epoch ECIECItodaytoday | Nov 9todaytoday | #76147.1Nemotron 3 Ultra | #1169.1best of the pickedGPT-6 Astra | #19159.3DeepSeek V4.1 Flash |
 ARC-AGI-2Scoretodaytoday | Nov 9todaytoday | #731.9%Nemotron 3.5 Super VL | #195.0%best of the pickedGPT-6 Astra | #1978.1%DeepSeek V4.1 Flash |
| GPQA DiamondAccuracytodaytoday | Nov 9todaytoday | #7585.4%Nemotron 3 Ultra | #295.8%best of the pickedGPT-6 Astra | #2193.2%DeepSeek V4.1 Flash |
 AA IntelligenceIndextodaytoday | Nov 9todaytoday | #3513.6Nemotron 3.5 Super VL | #356.9best of the pickedGPT-6 Astra | #1632.6DeepSeek V4.1 Flash |
 MMLU-ProAccuracy · staletodaytoday | Nov 9todaytoday | #7078.3%Nemotron 3 Nano 30b A3B | #389.3%best of the pickedo1 | #3385.0%DeepSeek-V3.2 |
 LiveBenchGlobal averagetodaytoday | Nov 9todaytoday | #5867.4%Nemotron 3 Ultra 550B A55B | #482.2%best of the pickedGPT-6 Astra | #881.1%DeepSeek V4.1 Flash |
| HLEAccuracytodaytoday | Nov 9todaytoday | | #154.8%GPT-6 Astra | |
| Coding4 | | | | |
|---|
 Aider PolyglotPass rate · staletodaytoday | Nov 9todaytoday | | #188.0%best of the pickedGPT-5 | #874.2%DeepSeek-V3.2 |
| WebDev ArenaArena ratingtodaytoday | Nov 9todaytoday | | #31,770best of the pickedGPT-6 Astra | #141,665DeepSeek V4.1 Flash |
 SWE-bench VerifiedResolved · staletodaytoday | Nov 9todaytoday | | #672.8%best of the pickedGPT-5.2 | #1170.0%DeepSeek-V3.2 |
| SWE-Bench ProResolvedtodaytoday | Nov 9todaytoday | | #490.2%GPT-6 Astra | |
| Agents and tools3 | | | | |
|---|
 BFCLOverall accuracy · staletodaytoday | Nov 9todaytoday | #8010.0%Llama 3.1 Nemotron Ultra 253b V1 | #763.0%best of the pickedo3 | #1256.7%DeepSeek-V3.2 |
| MCP AtlasPass ratetodaytoday | Nov 9todaytoday | #2963.1%Nemotron 3 Ultra | #1481.8%best of the pickedGPT-5.6 | |
| Search ArenaArena ratingtodaytoday | Nov 9todaytoday | | #11,255GPT-5.6 Sol | |
| Vision2 | | | | |
|---|
| Vision ArenaArena ratingtodaytoday | Nov 9todaytoday | #731,211Nvidia Nemotron 3.5 Super VL 120b A12B | #81,293best of the pickedGPT-6.1 Sol | #341,263DeepSeek V4.1 Flash |
| Document ArenaArena ratingtodaytoday | Nov 9todaytoday | | #51,486GPT-5.6 Sol | |
| Image2 | | | | |
|---|
| Text-to-ImageArena ratingtodaytoday | Nov 9todaytoday | #261,165Cosmos3 Super Text2image | #11,429best of the pickedGPT Image 2.5 Sunburst | |
| Image EditArena ratingtodaytoday | Nov 9todaytoday | | #11,517GPT Image 2.5 Sunburst | |
| Video1 | | | | |
|---|
| Text-to-VideoArena ratingtodaytoday | Nov 9todaytoday | | #131,367Sora 2 Pro | |
| Speech2 | | | | |
|---|
 Open ASRAverage WERtodaytoday | Nov 9todaytoday | #96.13 WERbest of the pickedNVIDIA-Nemotron-3.5-Transcribe-30B-A4B (nvidia) | #186.80 WERgpt-transcribe (openai) | |
| Speech ArenaArena ratingtodaytoday | Nov 9todaytoday | #591,064Magpie Multilingual 357m | #421,107best of the pickedTTS 1 Hd | |
| Embeddings1 | | | | |
|---|
| MTEBMean task scoretodaytoday | Nov 9todaytoday | #369.5%llama-embed-nemotron-8b (nvidia) | | |
| Speed1 | | | | |
|---|
| AA SpeedOutput speedtodaytoday | Nov 9todaytoday | #318 tok/sNemotron 3.5 Super VL | #1522 tok/sGPT-6.1 Sol | #1323 tok/sbest of the pickedDeepSeek V4 Pro 0813 |
| Price1 | | | | |
|---|
| AA PriceBlended pricetodaytoday | Nov 9todaytoday | | #6$2.00GPT-6.1 Terra | #3$0.53best of the pickedDeepSeek V4.1 Flash |
| Usage1 | | | | |
|---|
 OpenRouterWeekly tokenstodaytoday | Nov 9todaytoday | #143.0T tokNemotron 3 Ultra 550B A55B | #47.2T tokGPT-6 Luna | #156.0T tokbest of the pickedDeepSeek V4.1 Flash |
Green marks the best placing among the picked where two or more are ranked; ties are marked for each. Ranks are each board's own order (lower is better), scores as the board published them. Dash: not on that board.
Modelsby Glyph